混合专家论文趋势 2023–2026

主要结论

• 混合专家论文从 2023 年的 236 篇增长到 2026 年的 326 篇，三年间扩大到约 1.4 倍。
• NeurIPS 是发混合专家论文最多的会议，2023–2026 年共 354 篇，占四个会议总量的 39%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 4 篇增至 2026 年的 79 篇。
• 在「优化与缩放」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的缩放定律是 4.1 倍，混合专家是 1.4 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026A Biconvex Formulation for Stable Transport of Mixture Models with a Unique Solution
• ICML 2026$\phi$-Balancing for Mixture-of-Experts Training
• ICLR 2026A Federated Generalized Expectation-Maximization Algorithm for Mixture Models with an Unknown Number of Components
• CVPR 2026Divide Conquer and Aggregate Asymmetric Experts for Class-Imbalanced Semi-Supervised Medical
• NeurIPS 2025Advancing Expert Specialization for Better MoE
• ICML 2025$\texttt{I$^2$MoE}$: Interpretable Multimodal Interaction-aware Mixture-of-Experts
• ICLR 2025A Sanity Check for AI-generated Image Detection
• CVPR 2025Complexity Experts are Task-Discriminative Learners for Any Image Restoration
• NeurIPS 2024$\texttt{Model-GLUE}$: Democratized LLM Scaling for A Large Model Zoo in the Wild
• ICML 2024$\texttt{MoE-RBench}$: Towards Building Reliable Language Models with Sparse Mixture-of-Experts
• ICLR 2024Fusing Models with Complementary Expertise
• CVPR 2024MoDE: CLIP Data Experts via Clustering
• NeurIPS 2023[Re] Numerical influence of ReLU'(0) on backpropagation
• ICML 2023A Distribution Optimization Framework for Confidence Bounds of Risk Measures
• ICLR 2023Knowledge-in-Context: Towards Knowledgeable Semi-Parametric Language Models
• CVPR 2023Balanced Product of Calibrated Experts for Long-Tailed Recognition
• NeurIPS 2026AdERA: Adaptive Exponent Reuse for Lossless Allgather in Sharded MoE Training
• ICML 2026Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse
• ICLR 2026Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies
• CVPR 2026ERMoE Eigen-Reparameterized Mixture-of-Experts for Stable Routing and Interpretable Specialization

全部混合专家论文

共 59 篇，第 1 / 2 页

1. ICLR 2025A Sanity Check for AI-generated Image Detection
2. ICLR 2025Ada-K Routing: Boosting the Efficiency of MoE-based LLMs
3. ICLR 2025Balancing Act: Diversity and Consistency in Large Language Model Ensembles
4. ICLR 2025BEEM: Boosting Performance of Early Exit DNNs using Multi-Exit Classifiers as Experts
5. ICLR 2025CAMEx: Curvature-aware Merging of Experts
6. ICLR 2025Causal Order: The Key to Leveraging Imperfect Experts in Causal Inference
7. ICLR 2025Don't flatten, tokenize! Unlocking the key to SoftMoE's efficacy in deep RL
8. ICLR 2025Drop-Upcycling: Training Sparse Mixture of Experts with Partial Re-initialization
9. ICLR 2025Dynamic Mixture of Experts: An Auto-Tuning Approach for Efficient Transformer Models
10. ICLR 2025Dynamic Modeling of Patients, Modalities and Tasks via Multi-modal Multi-task Mixture of Experts
11. ICLR 2025Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders
12. ICLR 2025EC-DIT: Scaling Diffusion Transformers with Adaptive Expert-Choice Routing
13. ICLR 2025Efficient Dictionary Learning with Switch Sparse Autoencoders
14. ICLR 2025Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning
15. ICLR 2025Efficient Residual Learning with Mixture-of-Experts for Universal Dexterous Grasping
16. ICLR 2025Efficiently Democratizing Medical LLMs for 50 Languages via a Mixture of Language Family Experts
17. ICLR 2025ELBOing Stein: Variational Bayes with Stein Mixture Inference
18. ICLR 2025End-to-end Learning of Gaussian Mixture Priors for Diffusion Sampler
19. ICLR 2025Fiddler: CPU-GPU Orchestration for Fast Inference of Mixture-of-Experts Models
20. ICLR 2025Filtered not Mixed: Filtering-Based Online Gating for Mixture of Large Language Models
21. ICLR 2025Gaussian Mixture Counterfactual Generator
22. ICLR 2025Generalization Guarantees for Representation Learning via Data-Dependent Gaussian Mixture Priors
23. ICLR 2025GETS: Ensemble Temperature Scaling for Calibration in Graph Neural Networks
24. ICLR 2025Graph Sparsification via Mixture of Graphs
25. ICLR 2025HMoRA: Making LLMs More Effective with Hierarchical Mixture of LoRA Experts
26. ICLR 2025Intrinsic User-Centric Interpretability through Global Mixture of Experts
27. ICLR 2025Layerwise Recurrent Router for Mixture-of-Experts
28. ICLR 2025Many-Objective Multi-Solution Transport
29. ICLR 2025Mixture Compressor for Mixture-of-Experts LLMs Gains More
30. ICLR 2025Mixture of Attentions For Speculative Decoding
31. ICLR 2025Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
32. ICLR 2025Mixture of In-Context Prompters for Tabular PFNs
33. ICLR 2025Mixture of Parrots: Experts improve memorization more than reasoning
34. ICLR 2025Mixture-of-Agents Enhances Large Language Model Capabilities
35. ICLR 2025MoE++: Accelerating Mixture-of-Experts Methods with Zero-Computation Experts
36. ICLR 2025MoLEx: Mixture of Layer Experts for Fine-tuning with Sparse Upcycling
37. ICLR 2025Monet: Mixture of Monosemantic Experts for Transformers
38. ICLR 2025More Experts Than Galaxies: Conditionally-Overlapping Experts with Biologically-Inspired Fixed Routing
39. ICLR 2025MoS: Unleashing Parameter Efficiency of Low-Rank Adaptation with Mixture of Shards
40. ICLR 2025Multiple Heads are Better than One: Mixture of Modality Knowledge Experts for Entity Representation Learning
41. ICLR 2025NetMoE: Accelerating MoE Training through Dynamic Sample Placement
42. ICLR 2025No Need to Talk: Asynchronous Mixture of Language Models
43. ICLR 2025OLMoE: Open Mixture-of-Experts Language Models
44. ICLR 2025pMoE: Prompting Diverse Experts Together Wins More in Visual Adaptation
45. ICLR 2025Provable Uncertainty Decomposition via Higher-Order Calibration
46. ICLR 2025RegMix: Data Mixture as Regression for Language Model Pre-training
47. ICLR 2025ReMoE: Fully Differentiable Mixture-of-Experts with ReLU Routing
48. ICLR 2025Revisiting Prefix-tuning: Statistical Benefits of Reparameterization among Prompts
49. ICLR 2025Routing Experts: Learning to Route Dynamic Experts in Existing Multi-modal Large Language Models
50. ICLR 2025Self-MoE: Towards Compositional Large Language Models with Self-Specialized Experts

常见问题

2023–2026 年四大会议一共有多少篇混合专家论文？

共 904 篇，按年份分别是 2023 年 236 篇、2024 年 119 篇、2025 年 223 篇、2026 年 326 篇。

NeurIPS 2026 有多少篇混合专家论文？

NeurIPS 2026 有 75 篇混合专家论文，上一届是 86 篇。

ICLR 2026 有多少篇混合专家论文？

ICLR 2026 有 73 篇混合专家论文，上一届是 59 篇。

ICML 2026 有多少篇混合专家论文？

ICML 2026 有 132 篇混合专家论文，上一届是 57 篇。

CVPR 2026 有多少篇混合专家论文？

CVPR 2026 有 46 篇混合专家论文，上一届是 21 篇。

哪个会议的混合专家论文最多？

NeurIPS，2023–2026 年共 354 篇。